MathWorks Rolls Out AI-Enhanced Tools for Embedded Systems in Latest Release

MathWorks unveiled Release 2026a (R2026a) of its MATLAB and Simulink product families, bringing generative AI capabilities to embedded systems development. The latest iteration introduces Simulink Copilot to enhance Model-Based Design workflows and Polyspace Copilot to strengthen embedded software code analysis. These tools aim to boost engineering team productivity while preserving the rigor, traceability, and repeatability essential to complex system design.

The company is pursuing two parallel strategies for integrating AI into engineering work. It embeds copilots directly into familiar development environments—MATLAB Copilot, Simulink Copilot, and Polyspace Copilot—while simultaneously connecting MATLAB and Simulink capabilities to agentic workflows via MATLAB MCP Core Server and MATLAB Agentic Toolkit. This dual approach enables teams to grasp designs more rapidly, catch software defects sooner, and apply development and verification processes with greater consistency.

Avinash Nehemiah, Head of Product Management and Marketing for Design Automation at MathWorks, emphasized the importance of trustworthy AI in engineering contexts. "Engineering teams now have access to capabilities enabled by generative AI, and leaders need confidence that these translate into tangible engineering and business benefits," Nehemiah stated. "In engineering design and software verification, productivity improvements cannot come at the expense of rigor, traceability, or trust. MathWorks is committed to delivering grounded AI tools for engineering that help teams move faster while preserving the discipline and confidence required to develop complex engineered systems."

Simulink Copilot operates within the specific context of user models, established team processes, and MathWorks documentation. It generates model explanations, responds to queries about model behavior, and assists engineers in identifying relevant blocks and subsystems. The tool isolates problems, proposes solutions, and guides users through subsequent actions, accelerating design progress. Teams can leverage it to carry out standardized procedures that promote uniform development and verification practices.

R2026a introduces both Polyspace Copilot and Polyspace as You Code. Polyspace Copilot interprets static analysis results to help developers understand findings and resolve issues more swiftly. Polyspace as You Code allows developers to validate C and C++ coding standards and spot defects and vulnerabilities during development, including code generated with AI assistance. These capabilities help teams identify and address problems earlier while enhancing software quality throughout the development pipeline.

The Polyspace product family receives three key improvements. A new desktop application consolidates configuration and results management across the platform. Extensions to Polyspace Bug Finder introduce custom checkers and coding standards support. Polyspace Test gains software-sanitizing features for dynamic analysis of runtime errors. Collectively, these enhancements create a more integrated workflow for software quality across development, testing, and verification phases.

Beyond the AI-focused updates, R2026a delivers several enhancements across MATLAB and Simulink:

MATLAB Course Designer, a new offering, assists educators in building courses, courseware, labs, and assessments leveraging MATLAB and Simulink capabilities.

Simulink FMU Builder, also new, generates standalone Functional Mockup Units from Simulink models and C or C++ code, facilitating model exchange and integration activities.

MATLAB enhancements enable engineers to construct and distribute interactive webpages featuring visualizations without requiring a local MATLAB installation. The release also improves Python environment management and streamlines data transfer between MATLAB and Python.

Simulink improvements include streamlined access to frequently used functions through simplified, task-oriented context menus. The platform now supports simulating C and C++ code within models without language constraints or supplementary wrappers.

Wireless Network Toolbox allows engineers to model, simulate, analyze, and visualize wireless communication networks to assess complete system performance.

MATLAB Test gains the ability to generate starter tests, equivalence tests, and tests derived from command history using MATLAB Copilot. The toolbox also enables engineers to execute tests specific to the active file, minimizing unnecessary computational overhead.

Mapping Toolbox receives enhancements for geospatial analysis, including 3D building visualization, image overlay capabilities, and raster mapping features.

Signal Processing Toolbox introduces new Filter Designer and Filter Analyzer applications for digital filter design and analysis. Users can now label time-frequency data and extract signal characteristics using refined interactive capabilities.